Download A Comparative Analysis of Noun Incorporation Productivity in

A Comparative Analysis of Noun Incorporation
Productivity in English and German
Ian Westley
A thesis submitted in partial fulfillment of the requirements for
the degree of Bachelor of Arts in Linguistics
Swarthmore College
December 2016
Abstract
Noun incorporation (e.g. crab-walk or hand-hold) is defined by Feist (2013) as a
word or expression that includes a noun in the verbal constituent of an utterance. In this
paper, I compare the productivity of noun incorporation in English and German. My goal is
to determine whether the relationship between noun incorporation in English and German
is similar to that of nominal compounding. Berg et al. (2012) posit that German assigns a
more important role to compounding than English does, resulting in a higher occurrence of
nominal compounds in German. Specifically, their study finds that nominal compounds
composed of two constituents are more prominent and productive in German. Meyer
(1993) notes that nominal compound productivity in German has very few restrictions and
argues that picking up a German newspaper or magazine will expose the reader to a wide
range of novel noun-noun compounds. Since the expanding productivity of nominal
compounds in German is higher than in English, one might expect the same finding to be
reflected in noun incorporation. In this paper, I argue that such is not the case, and that
English makes more productive use of noun incorporation than German. I additionally
argue, along with Baker (2009) and Barrie & Mathieu (2016) that one shouldn't expect the
productivity of English and German noun incorporation to perfectly mirror Berg et al.'s
findings, since noun incorporation is a syntactic process whereas compounding is a
morphological process. In order to facilitate this study, I adopt Mithun's (1984)
classification of noun incorporation, in which she breaks the process of noun incorporation
down into four categories. Based on the existence (or lack thereof) of these four types of
noun incorporation in German and English, I conclude that English noun incorporation
displays comparatively superior productivity. To carry out this comparison, I cite examples
from a variety of studies (see Feist 2013; Barrie 2011; Barrie & Spreng 2009; Barrie &
Mathieu 2016) and additionally interview two native German speakers.
*1 would like to thank Ted Fernald and Emily Gasser for their immensely helpful remarks on
previous drafts of this thesis. Additionally, I would like to thank Nathan Sanders for his
assistance with my associated squib, and for teaching me the proper citation format. I would
also like to thank Ute Bettray and Peter Baumann for allowing me to interview them and
gather data from our discussions. Finally, I would like to thank the many linguistics majors
who served as peer editors and generally provided useful feedback as my thesis progressed.
2
Table of Contents
1 Introduction
4
2 Compounding in English and German
5
2.1 English Nominal Compounds . . . . . . . . . . . . . . . . . . . . . . 6
2.2 German Nominal Compounds . . . . . . . . . . . . . . . . . . . . . . 8
2.3 Thomas Berg et al.'s Discussion of Nominal Compounds ..... 9
3 Is Noun Incorporation a Type of Compounding?
12
4
16
Mithun's Theory of Noun Incorporation
5 Applying Mithun's Typology to Greenlandic Inuit, English, and
German
22
5.1 Noun Incorporation in a Polysynthetic Language. . . . .
22
5.2 Noun Incorporation in English. . . . . . . . . . . . . . . . . . . . 24
5.3 Noun Incorporation in German . . . . . . . . . . . . . . . . . . . . 28
6 Native German Speakers
31
7 Conclusion
37
Appendix A
39
References
40
3
1 Introduction
At the most basic level, noun incorporation consists of a process in which "a noun
stem binds to a verb stem ... in order to produce a derivative verb stem" (Rice &
Prideaux 1991:283). The study of noun incorporation in language dates back over a
century, to Edward Sapir's article in 1911 documenting noun incorporation in
Native American languages. In this paper, I compare noun incorporation in English
and German in an attempt to contrast the productivity of this process in two closely
related West Germanic languages.
Both German and English make extensive use of lexical compounding, with
German offering a higher frequency of lexical compounds than its West Germanic
cousin. Thus, the goal of this paper is to determine whether the same relationship
found between compound nouns in German and English (specifically, higher
productivity in German nominal compounding) holds true with incorporated nouns
in the same languages. As closely related languages in the same language family, one
might suspect the syntax and semantics of a given linguistic process, such as noun
incorporation, to remain fairly similar. In order to facilitate this study, I utilize
Marianne Mithun's four types of noun incorporation, proposed in 1984.
As a general overview, Mithun identified four distinct types of noun
incorporation that occur in the world's languages, distinguishable by various
syntactic and semantic traits. These four types operate in an implicational hierarchy,
such that a given language can't utilize Type 4 noun incorporation unless it also
utilizes Types 1-3. This platform provides a strong initial base for comparison, as a
difference in the number of types found in German versus in English, or in the
4
number of instances of each type, would hint at a variation in the productivity of
each language's noun incorporation process.
It is worth noting briefly that the derived verb stem resulting from noun
incorporation can occasionally be altered to produce a derived noun stem as well,
e.g. to babysit and babysitter (although this may actually be an example of backformation from babysitter to babysit; see Feist 2013:162). When I refer to noun
incorporation throughout this paper, I exclusively imply incorporated nouns that
surface as derived verb stems (or gerunds), and from this point forward I analyze
the process of noun incorporation as one that produces a derived verb stem.
In the following section, I will briefly discuss compounding in English and
German. In Section 3, I discuss whether noun incorporation is a subsection oflexical
compounding, or if it is an entirely unrelated phenomenon. In Section 4, I give a
summary of Mithun's noun incorporation theory. In Section 5, I apply Mithun's
theory to my sample of data and identify any potential outliers of Mithun's typology.
I additionally provide a brief overview of noun incorporation in a polysynthetic
language (Greenlandic Inuit). In Section 6, I present data found upon interviewing
two native German speakers. In Section 7, I make a final comparison between the
noun incorporation productivity of these two languages, and I conclude with my
results.
2
Compounding in English and German
Berg et al. (2012) provide a direct and relevant comparison of English and German,
in which they perform a comparative study regarding the productivity, complexity,
and frequency of nominal compounding in English vs. German. They argue that
5
nominal compounding occurs more frequently, and is more productive, in German
than in English. Berg et al. remark, "the German system accommodates compounds
more easily than the English one because languages favor morphological processes
which create syllabic templates which already exist for base words" (Berg et al.
2012:293). In Section 2.3, I further explore their paper, and provide additional
insight regarding Berg et al.'s argument. This paper attempts to prove whether the
same relationship described by Berg and his colleagues (with regards to
productivity) holds true between noun incorporation in German and in English.
2.1 English Nominal Compounds
Compound words can be defined as lexemes in which "their internal structure
shows two or more lexemic bases ... forms which in other places in the language
inflect independently and can on their own act as the heads of relevant phrases"
(Bauer 2006:485). In English, nominal compounding (compounding noun stems)
exists as a highly productive process, although there are differing lexical restrictions
on its use when compared with German nominal compounding. Additionally,
English compounds often appear as two individual written words, even though they
act as one lexeme (e.g. table tennis).
Lexicalization also frequently occurs in English nominal compounds,
occasionally to the point where the meaning of a compound word can't be entirely
derived from its lexical morphemes (e.g. tongue twister). Thus, many English
compound nouns hold a distinct lexical entry from the lexical morphemes that they
contain, referred to by Berton et al. (1996) as 'semantic compounds.' For the
remainder of this paper, I will refer to such compounds as 'lexicalized compounds,'
6
as I find the term 'semantic compound' to be inadequate in describing the specific
characteristics of this compound type. In lexicalized compounds, "the semantic
content of the compound cannot be derived from the semantic content of its
constituent parts" (Berton et al. 1996:1165). As a result of these lexicalized
compounds, the productivity of nominal compounding is hampered. For example,
while blueberry is a perfectly valid lexicalized compound in English, *greenberry
would not be a possible lexicalized compound, since the word greenberry has not
been lexicalized by (most) native English speakers to contain a specialized meaning
that is not simply the sum ofits constituents. It is worth noting, however, that
greenberry may still potentially appear as a compound in English in which its
meaning is directly derived from the constituents green and berry, thus referring to
a berry that is green (rather than a specific species of berry, as seen with blueberry).
The opposite of lexicalized compounds are referred to by Berton et al. as
'orthographic compounds,' in which "the constituent parts retain their original
semantic content, e.g. apple tree [or pear tree]" (Berton et al. 1996:1165). For the
remainder of this paper I will refer to such compounds as 'compositional
compounds,' since their semantic interpretation exactly reflects the meaning of the
compound's individual constituents. While compositional compounds certainly
occur in English, they are not as easily identifiable in writing since English examples
are often written in two distinct words (e.g. NBC broadcast). Berg et al. explicitly
state this disparity, remarking, "whereas English compounds are usually written in
two words, German compounds are typically written together (contrast Tischtennis
with table tennis)" (Berg et al. 2012:276). As Berg et al. additionally note, the
7
difficulty in identifying English nominal compounds stretches far beyond simple
word boundaries, as morphology further enhances the difficulty of accurate
compound identification. "English is notorious for its difficulty of distinguishing
compounds from phrases. Neither spelling nor stress placement serve as reliable
criteria for compound identification" (Berg et al. 2012:277). For example, while
English compounds are usually stressed on their initial element (e.g. hedgehog,
fastball), there are also a bevy of English compounds that defy this trend (e.g.
Madison Avenue, Boston Marathon) (Plag et al. 2008:761). As a whole, compounding
is a highly productive process in English, despite occasionally being difficult to
identify at first glance.
2.2
German Nominal Compounds
In German nominal compounds, morphemes of the same semantic role may be
substituted for one another much more freely than in English. Consider the
following examples of German compositional compounds:
1) Giasf/achenreinigung
Glass-surface-cleaning
Glass surface cleaning
2) Unterhaitsreinigung
Floormat-cleaning
Floormat cleaning
The initial morphemes in a German compound word, referred to in German
as determinative elements, can be substituted for one another easily, as seen in
sentences (1) and (2). In German, the heightened productivity of nominal
compounding is a direct result of the overwhelming possible iterations of
grammatically acceptable compositional compounds. As Meyer writes, "each issue of
8
any newspaper or magazine contains a whole range of novel NN -compounds. Hence
the reader is not bound to have read them, but is able to understand immediately
within the context of the corresponding text" (Meyer 1993:1). It would therefore
appear that the amount of possible German compositional compound nouns
numbers in the billions, with hundreds of thousands oflexical morphemes
(nominals) able to substitute for one another, so long as they exist within the same
semantic role. As Meyer readily admits, "German is so productive in forming ad-hoc
compounds that there seem to be no restrictions at all" (Meyer 2013:1).
This finding is reinforced by Berg et aI., who claim that "German two-noun
compounds [the large majority of German nominal compounds 1make a larger
contribution to the formation of new nouns than do English compounds of the same
size" (Berg et al. 2012:290).
2.3
Thomas Berg et al.'s Discussion of Nominal Compounds
In comparing the nominal compounding structures of German and English, Berg and
his colleagues specifically aimed to study three aspects: frequency, productivity, and
complexity of compound nouns within the two languages. By researching these
three measures of compounding, the goal of their work is to study the relationship
between these measures, e.g. not only what the productivity of nominal
compounding looks like in German, but also how increased productivity effects the
frequency of compounding.
In order to facilitate this study, Berg and his colleagues consulted closely
related pieces of literature in German and English: contemporary political
magazines. An additional study was conducted with German and English fictional
9
novels, with the linguists selecting a specified (and constant) number of words from
each publication, to keep the sample sizes from each language and corpus as
identical as possible. After navigating through a number of difficulties in adeptly
identifying English nominal compounds, Berg et al. provide data for their three
measures, juxtaposing each measurement in English and German.
Berg et al.'s findings concerning the frequency of nominal compounding in
German and English confirmed pre-existing linguistic work that compound nouns
appear more frequently in German literature than in English literature. The medium
of publication (political science journal versus fictional novel) seemed to have a
minimal impact on the prominence of nominal compounds. The following table is
taken directly from Berg et al.'s publication (Berg et al. 2012:280):
Table I
T)-pefreql/ency of allnOl/ns (N
=
41.068)
A: Journalese (N = 26,378)
B: Fiction (N = 14,690)
Language
noncompound
nouns
compounds
lolal
noncompound
nouns
compounds
lolal
German
English
GtE ratio
6724
6738
10
8661
4255
2.0
15,385
10,993
14
4909
4537
11
3467
1777
2.0
8376
6314
L3
Using this data, Berg and his colleagues proceeded to measure the effect that
heightened compound complexity has on frequency, and found that an inverse
relationship exists in both languages. As can be seen in the table below, as
complexity increases in both German and English nominal compounds, frequency
decreases. A more substantive finding, however, arises by examining the ratio at
which German nominal compounds of a given complexity occur compared to English
compounds of the same complexity. While simple compound nominals (i.e.
10
containing rno or three constituents) occur much more frequently in German than
in English, the observed frequencies gradually converge when more complex
compounds are encountered, so that compound nouns with greater than three
constituents appear slightly more frequently in English literature than in German
(Berg et al. 2012 :281):
Table 3.
Tjlpe frequency of all compollnds as a fllllClioll of complexity
A : Journalese
Complexity:
two-member
three-member
four-member
five-member
Total
Language
German
English
G/E ratio
7837
3648
21
770
549
1.4
52
56
0.9
2
2
1.0
4255
8661
B: Fiction
Complexity:
two-member
three-member
four-member
five-member
Total
Language'
German
English
G/E ratio
3267
1604
20
196
166
4
7
06
0
0
3467
1777
12
Finally, in order to measure productivity of nominal compounds in German
and English, Berg and his colleague had to eliminate all redundancies in the data,
narrowing the list of available nominal compounds to "hapax legomina i.e., items
occurring only once in a given data set" (Berg et al. 2012:287) . By reducing the data
to nominal compound hapaxes, and subsequently dividing this quantity by the
number of nominal hapaxes, Berg et al. were able to quantify the productivity of
nominal compounding in both languages (P*). As the measures of productivity
below show, Berg and his colleagues were able to reaffirm that German nominal
11
compounds (particularly tvvo-member compounds) appear in more instances than
the same process does in English (Berg et al. 2012:289):
Table 8.
Productivity mllles (P*) for compollllds as afimclioll ofcomplexity
A: Journalese
Complexity:
two-member
three-member
four-member
Language:
Gennan
English
0.67
0.51
0. 18
0. 15
0.02
0.02
B: Fiction
Complexity:
two-member
three-member
four-member
Language:
Gennan
English
0.53
0.38
0.07
0.07
0.00
0.00
A conceivable expectation based on the data above is that the productivity of
German vs. English noun incorporation will closely mirror the trends of nominal
compounding. One might expect German to be the more productive language in
incorporating nouns, especially when concerning incorporated nouns that aren't
complex. However, this expectation is not as justifiable as one might imagine based
on the apparent similarities betvveen noun incorporation and nominal
compounding.
3 Is Noun Incorporation a Type of Compounding?
A looming question arises in what can be safely inferred from Berg et al.'s paper and
applied to our expectations of noun incorporation in German and English. At the
heart of this uncertainty lies an even larger question that must be answered namely, what is noun incorporation in and of itself? Is noun incorporation simply
verbal compounding. or is a wholly different process occurring? For one can't make
12
assumptions based on Berg et al: s work without first understanding how noun
incorporation, as a process, relates to compounding. Such is the question that Barrie
& Mathieu (2016) attempt to answer. Barrie and Mathieu argue that noun
incorporation is not an example of compounding. As evidence, they provide the
following dataset, composed of examples of noun incorporation utilizing discourse
reference in the polysynthetic languages of Mohawk and Ojibwe (Barrie & Mathieu
2016:3):
(3)
ThetA're' wa' -ke-nakt-a-hninu-'
[Mohawk]
yesterday FACT-I.SG .AG-bed-0 -buy-PUNC
I-k-her-e'
Uwari A-ye-nuhwe'-ne'
0 -lsG.AG-think-STAT Mary FUT-3.F.SG.AG-like-PUNc
'I bought a bed yesterday. 1 think that Mary will like it.' (Baker 1996:288)
(4)
ngii-moonahapnii mii dash ngii-giziihiiginigan
n- gii- moonah
-apnii -e
mii dash
1- PST- dig
-potato -VAl and then
n- gii- giziibiiginig-an
-3PL
1 PST- wash
'I dug up potatoes, and then 1 washed them.' (BI, 2008-12-17)
[Ojibwe]
The possible implementation of discourse reference regarding incorporated
nouns constitutes a primary piece of evidence for Barrie and Mathieu (2016) - in
some instances of noun incorporation, the incorporated noun can be referred back
to later in the same iteration, while such is not the case with compounds. For
example, the following English sentence would be ungrammatical as a result of this
rule: 'The teethmarks were especially sharp because afthem (where them refers to
the teeth). Sentences (3) and (4) illustrate the grammatical inclusion of discourse
reference in polysynthetic noun incorporation. In sentence (3), nakt ('bed') is
referred to later in the same phrase, while in sentence (4) the same process occurs
with apnii ('potatoes')' Discourse reference is one of two main pieces of evidence
13
cited by Barrie and Mathieu to differentiate noun incorporation from compounding
(with the other being adjunct stranding). I highlight this particular distinction
because it illumines our discussion of noun incorporation, and thus resurfaces in
Section 5.2.
The explanation put forth by Barrie and Mathieu (2016) to describe the
disassociation between compounding and noun incorporation is to label noun
incorporation as a syntactic phenomenon, as opposed to compounding, which is
considered a morphological (or perhaps morpho syntactic) process. Additional
support for this argument lies in the fact that initial noun stems in English nominal
compounds are almost impossible to modify, and thus the two constituents of an
English nominal compound are virtually never syntactically independent. Note that
there are exceptions to this statement (for example, toothmarks and teethmarks are
both grammatical compounds in English, as are passer-by and passers-by). These are,
however, exceptions to the overarching pattern that initial noun stems in English
nominal compounds cannot be modified. Barrie and Mathieu remark on this lack of
syntactic independence, arguing, "we take this contrast to show that noun
incorporation is undeniably syntactic in comparison with English compound
formation ... we take compounding as a case of root-root merger in the lexicon"
(Barrie & Mathieu 2016:4).
Mark Baker reinforces the argument that noun incorporation is a syntactic
process in many additional publications (see Baker 1988, 1996, 2003, 2009). Baker
argues that noun incorporation is a syntactic phenomenon as opposed to a
morphological or lexical phenomenon, positing that instances of noun incorporation
14
occur due to head movement. A simplified representation of this head movement
can be observed in the following syntactic tree diagram (Baker 1996:13):
(9)
-------------s
VP
NP
V
_______
N
V
I
I
tobacco buy
NP
I
N
I
(!obacco)
In the preceding tree diagram, Baker depicts that the underlying
representation of the incorporated noun 'tobacco-buy' doesn't involve noun
incorporation, and thus appears as the VP 'buy tobacco.' However, upon reaching
surface form, the head of the NP phrase, 'tobacco,' is fronted to form the
incorporated noun 'tobacco-buy.'
Barrie and Mathieu comment on Baker's analysis, claiming that this
representation is an "elegant solution that captures many properties associated
with noun incorporation and it is a prevalent account of noun incorporation" (Barrie
& Mathieu 2016:5). While I accept Baker's claim that noun incorporation is a
syntactic phenomenon, and thus an unrelated process to lexical compounding, I find
Barrie & Mathieu's evidence inconclusive with specific regard to English and
German noun incorporation. We do not see either discourse reference or adjunct
stranding (the &0 pieces of evidence that Barrie and Mathieu use to vvarrant a
syntactic view of noun incorporation) in either English or German noun
incorporation. Yet, as we see in Sections 5.2 and 5.3, both English and German
contain processes of noun incorporation that fit into Mithun's original typology.
Therefore, I understood Barrie and Mathieu's introduction of discourse reference
15
and adjunct stranding not to indicate that such processes are universally obligatory
features of noun incorporation, but rather that their presence in certain languages
suggests that noun incorporation is a syntactic process. While discourse reference
and adjunct stranding don't occur in the noun incorporation of English or German,
and thus don't assist in our differentiation between noun incorporation and
compounding in these specific languages, the fact that these processes can occur in
noun incorporation fundamentally contrasts it with compounding, in which such
structures aren't possible.
In this paper, I adopt Baker's proposal that noun incorporation is an example
of head movement that occurs between the underlying representation and surface
structure of a given word, thereby further differentiating it from compounding.
Certainly, if one is to accept the findings that noun incorporation and lexical
compounding are two separate processes, then it is illogical to conclude that the
relationship between noun incorporation in German and English will exactly mirror
the findings of Berg et al. regarding nominal compounding. Thus, if the productivity
(or frequency) of noun incorporation in English were to surpass that of German, it
shouldn't be terribly surprising, since the relations argued by Berg and his team
can't be expected to map directly to an entirely different, syntactic, process.
4 Mithun's Theory of Noun Incorporation
Mithun (1984) highlights four distincttypes of noun incorporation found in the
world's polysynthetic languages. Mithun includes solely examples from
polysynthetic languages because it was widely believed, even by prominent linguists
such as Mark Baker, that "only polysynthetic languages can correctly be considered
16
to exhibit true noun incorporation" (Massam 2009: 1078). Thus, at the time of her
paper's publishing, there were no examples provided in Germanic languages
because it was conventional theory that no such examples existed. With regard to
such examples, C. Weggelaar writes that noun incorporation "occurs in Frisian, but,
as far as I know, not in other Germanic languages" (Weggelaar 1986:301). While it
has since been shown that noun incorporation occurs in both German and English
(Feist 2013; Barrie & Spreng 2009; Barrie 2011; see also Massam 2009), I mention
this note on polysynthetic languages because many of my examples used below to
illustrate Mithun's theory occur in polysynthetic languages.
The syntactic variations in incorporated nouns alluded to above are
primarily reliant on whether the derived verb stem is transitive or intransitive. The
semantic differences typically arise from the purpose of the incorporated noun stem
- for example, whether it acts as a classifier (see examples 6 and 7), or whether it
serves as a direct vs. indirect participant (see Feist 2013). Feist (2013) perceives the
distinction between direct and indirect participation, with regards to the
incorporated noun stem, as essential to his typology of English noun incorporation.
Thus, words such as mountain-climb, in which the complement of the verb, and
therefore the direct participant, is incorporated, occupy a different category than
words like filterJeed, in which the adjunct, and thus the indirect participant, is
incorporated. For the purposes of this paper, however, I will rely on Mithun's
typology, and therefore will not distinguish between direct and indirect participants
in my comparisons of German and English noun incorporation.
17
Type 1 noun incorporation is the least complex example of noun
incorporation - a form of lexical binding in which a noun stem and verb stem
combine to create a newly derived verb stem. As Massam writes, Type 1 noun
incorporation occurs when "a verb incorporates its object to form an intransitive
compound verb" (Massam 2009: 1078). Mithun (1984) expands on this description,
arguing that Type 1 noun incorporation denotes an institutionalized, non-specific
activity or state. Mithun's original definition of Type 1 noun incorporation
additionally allows for theta roles that aren't objects (e.g. instruments) to be
incorporated. An important observation to be made is that semantics plays a
relatively small role in the definition of Type 1 noun incorporation; rather, syntax
accounts for most of Type l's defining features. The following example contrasts a
sentence that doesn't utilize Type 1 noun incorporation (3a) with one that does
(3b). Both sentences have virtually identical meaning and occur in Mokilese, an
Austronesian language:
3)
aJ
Ngoah kohkoa oaring-kai.
1
grind coconut-these
'I am grinding these coconuts.'
bJ Ngoah ko
oaring.
1
grind coconut
'I am coconut-grinding.'
(Massam 1984:849)
Nate that while ko oaring ('coconut-grind') hasn't been bound
orthographically into a single freestanding word, it nevertheless appears
syntactically as one bound unit. The resulting incorporated noun isn't lexicalized, as
the underlying meaning of the statement isn't altered from (3a) to (3b).
18
Type 2 noun incorporation is structurally similar to Type 1, except that "the
verb remains transitive, and a word denoting a peripheral argument takes the
position (and case) of the incorporated word" (Feist 2013:160). Type 2 noun
incorporation is therefore labeled as the manipulation of case, as case manipulation
is one of the features central to the understanding of Type 2. Additionally, the first
instance of a significant semantic variation (within Mithun's typology) appears in
constricting the meaning of the incorporated noun stem, as well as the novel direct
object. Consider the following example of Type 2 noun incorporation from Yucatec
Mayan:
4) k-in-c'ak-ce-t-ik
in-kool
INCOMP- I-chop-tree-TR- IMP my-cornfield
'I clear my cornfield.'
(Mithun 1984:858)
In (4), the verb remains transitive by "allowing for another oblique argument ...
to become a derived object," in this case the 'cornfield' (Massam 2009:1078-9). The
incorporated noun stem 'tree' primes the peripheral argument 'cornfield,' as the
literal translation of (4) would more closely resemble 'I chop down my cornfield as
one would chop down a tree.' Since the incorporated noun stem 'tree' is modifying
the verb stem 'to chop,' it allows for the previously oblique argument 'cornfield' to
become a derived object of the verb 'to tree-chop.'
Type 3 noun incorporation can syntactically consist of either Types 1 or 2, but
the incorporated noun is used "to manipulate discourse structure ... a previously
discussed noun is backgrounded as old [or previously known1information"
(Massam 2009:1079). Consider the following conversation in Hautla Nahuatl, in
which Speaker B uses Type 3 noun incorporation:
19
5)
Speaker A) askeman ti-'-kwa nakatl
never
you-it-eat meat
'You never eat meat.'
Speaker B) na' ipanima ni-naka-kwa
I always I-meat-eat
'I eat it (meat) all the time.'
(Mithun 1984:860)
While the transitivity of the derived verb stem isn't significant in identifying
Type 3, the semantic input of the incorporated noun stem is. As Feist notes, "the
backgrounding [in Type 3 noun incorporation 1resembles the effect of using a
pronoun in English - 'J eat it all the time.'" (Feist 2013:161).
Finally, Type 4 noun incorporation requires the verb to remain transitive (and
thus is also an instance of Type 2) with the added wrinkle that noun classification
must occur. More specifically, in Type 4 the incorporated nominal stem must act as
the classifier for the newly derived object. As Massam writes of Type 4 noun
incorporation, "the incorporated nominal is less specific, and provides the class of
the object" (Massam 2009:1079). Consider the following examples in Chamorro (6)
and Gunwinggu (7):
6) Giii-[ga]
si
Jose kiitu
AGR.have-pet UNM Jose cat
'Jose has a pet cat.'
(Chung & Ladusaw 2004:120)
7) bene-dulg-naIJ
man,qaralaljmayn
they.two-tree-saw cashew.nut
'They saw a cashew tree.'
(Mithun 1984:867)
In sentences (6) and (7), note the semantic relationship between the
incorporated noun and the newly derived object. In what is referred to as 'the
hierarchy of class inclusion,' Chung and Ladusaw remark that "either an antecedent
can be subordinate to its discourse anaphor on the hierarchy, or the two can be
20
coextensional" (Chung and Ladusaw 2004:119). In (6), cat is subordinate to pet,
while cashew is subordinate to tree in (7). Thus, it would be grammatically incorrect
to switch the noun stems in (6), so as to say that *Jose cat-has a pet. Chung and
Ladusaw's observation aligns with Massam's, as it would be difficult to imagine an
inferior incorporated noun stem (in terms of the hierarchy of class inclusion)
assigning the class of a hierarchically superior object.
While I have covered each type of noun incorporation as proposed by Mithun, it
must be noted that this list is perhaps not wholly comprehensive of all forms of
noun incorporation occurring in the world's languages. As this theory was proposed
in 1984, more contemporary papers (see Feist 2013) have argued that Mithun's
theory is too restrictive, and that certain examples don't fit perfectly within the
boundaries of Mithun's 4 types. For example, Feist designs his own, individual,
typology specifically aimed to identify the various types of noun incorporation
found in English. While his typology is firmly based on Mithun's theory, Feist adds to
her typology by making a further semantic distinction between direct and indirect
participants (in terms of the incorporated noun stem). In doing so, however, Feist
provides insights as to what each of his 'types' of noun incorporation would be
considered under Mithun's typology. In this paper, the purpose of including
Mithun's theory is not to identify every individual form of noun incorporation
occurring in English and/or German, as Feist sets out to achieve. Rather, I use
Mithun's theory as a template to identify what (if any) significant differences exist
between the productivity of noun incorporation in English vs. German.
21
5 Applying Mithun's Typology to Greenlandic Inuit, English and German
In this section, I apply Mithun's typology to three languages - Greenlandic Inuit,
English, and German. For the three languages, I isolate and analyze each possible
type of noun incorporation that may occur, gradually ascending the inherent
hierarchy proposed by Mithun. I first analyze the polysynthetic language of
Greenlandic Inuit in order to introduce a language with an extremely productive
noun incorporation process. I then carry out a similar analysis in both English and
German, depicting each type of noun incorporation that can occur in these two West
Germanic languages.
5.1 Noun Incorporation in a Polysynthetic Language
The most productive and extensive usage of noun incorporation is found in
polysynthetic languages, in which morphemes can be easily bound together and
thus the limit of possible morpheme combinations stretches far beyond what can be
done in languages such as English or German. Before analyzing English and German,
examining noun incorporation in a polysynthetic language such as Greenlandic Inuit
puts our later analysis into perspective. Furthermore, noun incorporation in
Greenlandic Inuit can be rendered into every type described in Mithun's theory.
However, even in polysynthetic languages such as Greenlandic Inuit, there
are still restrictions on the constituents of an incorporated noun. For example, there
is a limited list of verb stems that may grammatically incorporate a noun stem, even
though there are few, if any, restrictions on the noun stem that is incorporated. Verb
stems such as 'be: 'go: 'have: and 'eat' are commonly found in Greenlandic Inuit
noun incorporation. As Sherkina-Lieber and Murasugi write, "only a restricted set of
22
bound affixal verbs can incorporate nouns, and incorporation is obligatory with
those verbs" (Sherkina-Lieber & Murasugi 2015:2). This is a trend observed
throughout the noun incorporation of many languages (although other
polysynthetic languages, such as Mohawk, impose no such constraints on what verb
stems can and can't incorporate noun stems). In the agglutinative language of
Chamorro, for example, the only verb stem that can grammatically incorporate a
noun stem is the verb 'to be' (see Chung & Ladusaw 2004). Despite this apparent
constraint, Greenlandic Inuit makes extensive use of noun incorporation, and Types
1-4 are all exhibited.
Type 1 noun incorporation is the most common rendition that appears in
Greenlandic Inuit. Consider the following sentence, in which the verb stem taaq
('get') incorporates its direct object, saa ('table'):
8) Ilisaiji
saa-taaq-tuq
uujuajar-mik
Teacher.ABS table-get-PART.3S green-MIK
(Sherkina-Lieber & Murasugi 2015:2)
'The teacher got a green table.'
Note that due to the aforementioned restriction in which only certain verb
stems may incorporate a noun stem, replacing the verb 'to get' with 'to buy' in (8)
would be ungrammatical.
Type 2 noun incorporation also exists in Greenlandic Inuit, although the
morpheme order differs from examples later shown in English. Consider the
following sentence in Kalaallisut, the most commonly spoken dialect in Greenland of
Greenlandic Inuit:
9) Tutt-up
neqi-tor-punga.
Caribou-REL meat-eat-INDIC.1S
'I ate reindeer meat.'
(Johns 2009:187)
23
Note that sentence (9) is likely also an example of Type 4 noun incorporation,
since the type of meat being eaten is reindeer meat. Thus, the noun steam 'meat' acts
as a classifier by priming the noun stem 'caribou.'
Finally, Type 3 noun incorporation can also be found in the Kalaallisut dialect, as
is exhibited in the following sentence:
10)Suulut pappiara-mik sana-a-mik
nasa-qar-poq.
Suulut paper-MIK make-PASS-MIK hat-have-INDIC.3S
'Suulut has a hat made of paper.'
(Johns 2009:186)
In (10), the sentence includes Type 3 noun incorporation because the 'paper' is
simply background information describing what the 'hat' is made of. Thus,
Greenlandic Inuit productively utilizes Types 1-4 of noun incorporation under
Mithun's typology. While we find that neither English nor German can produce all 4
types of noun incorporation, Greenlandic Inuit can be used as a basis of comparison
by which to measure the productivity of noun incorporation in these West Germanic
languages.
5.2 Noun Incorporation in English
As Feist (2013) writes, "productivity is taken to denote the variety of forms
incorporation takes, and the number of instances of each form" (Feist 2013:177). In
English, noun incorporation productively utilizes Types 1-3 of noun incorporation.
In this subsection, I observe how each type appears in English, and the productivity
of each type.
Type 1 noun incorporation, being the simplest rendition of noun
incorporation, occurs frequently in the English language and is a fairly productive
process. When I attempt to come up with an example of noun incorporation, Type 1
24
incorporated nouns come to mind most easily. For example, consider the
incorporated nouns to mountain-climb, to sleep-deprive, or to berry-pick. In all three
prior examples, the verb is intransitive (none of the three verb stems can adopt a
direct object) and the verb stems have all incorporated their original direct objects
to derive a novel verb stem. Type 1 noun incorporation is highly productive in
English, especially in conversational context. Consider the following example, in
which berry-pick is used to derive a new incorporated noun:
11 ) While John likes to berry-pick on his farm, Mary prefers to lime-pick.
While lime-pick isn't likely to appear in one's pre-existing lexicon, conversational
context (initially establishing the lexicalized incorporated noun berry-pick) makes
lime-pick understandable and grammatically acceptable in (11). Note that the same
process can be applied to nominal compounding in English, such that blueberry can
likely be included in an utterance to prime the grammaticality and understanding of
green berry.
Type 2 noun incorporation appears to be far less productive in forming new
words of English (when compared to Type 1), but is certainly present. Consider the
following example, a slightly altered version of the sentence found in Feist
(2013:170):
12)The flight attendant's job was to help busy travelers catch their connecting flights,
or kindly hand-hold them through their delays.
Nate that the incorporated noun remains transitive in (12), with the
incorporated noun stem hand denoting the peripheral argument them (i.e. the
travelers whose flights have been delayed). However, this example is lexicalized,
25
and there are also a number of compositional Type 2 incorporated nouns such that
the meaning of the derived verb stem is a sum of the two constituents. Additional
examples of Type 2 noun incorporation in English include babysit, GPS-track,finger-
paint and test-drive. Although not discussed in this paper, an idea for future research
would be to study the productivity of each individual type of noun incorporation in a
given (polysynthetic) language, to see if productivity decreases as one ascends the
hierarchical structure of Mithun's typology.
Finally, Type 3 noun incorporation also occurs in English, usually in the form of
an intransitive verb (Type 1). Consider the following example of Type 3, in which
the incorporated noun stem exists to background incidental information:
13) Manta rays filter-feed in the murky broth of plankton.
(Feist 2013:168)
In (13), the sentence literally means that the manta rays are feeding on the
plankton using their filters, such that the filter becomes incidental information that
acts as an indirect participant in the sentence. This is additionally distinct from
many English instances of Type 1 in that the verb is no longer incorporating its
initial direct object [filter is not the incorporated object of feed). Once again, Type 3
occurs frequently in English, as the newly derived verb stem is intransitive, but the
productivity is hampered by the lack of possible constituent combinations. In other
words, English noun incorporation is restricted by the rule that a specific verb stem
can only incorporate a limited number of noun stems.
While the presence of three of Mithun's types indicates that noun incorporation
in English is at least partially productive, the relative dearth of compositional
incorporated nouns in English indicates that productivity is relatively lacking when
26
compared to the productivity of English nominal compounds. For example, a highly
productive means by which to create compositional compounds in English is to
think of street names. Madison Avenue, Broad Street, and Baltimore Pike are all
nominal compounds in English, in which thousands of noun stems may appear
directly in front of the words street or avenue. In English noun incorporation,
however, verb stems are highly limited in the noun stems that they can feasibly
incorporate. Feist (2013) argues that "different forms of the verb are productive to
very different extents ... uses of [finite verbs and] the past tense are uncommon,"
while a large majority of noun incorporation instances are nonfinite (Feist
2013 :178). Additionally, Feist notes that certain noun and verb stems are more
prominent in English noun incorporation than others, such as -walk, hand-, and -eat.
A conceivable predicament that this analogy of English noun incorporation
encounters is one of the arguments presented by Barrie and Mathieu (2016:3) in
describing the disparities between noun incorporation and nominal compounding.
In doing so, the authors argue that noun incorporation is a syntactic process
whereas compounding is a morphological process. A key reason cited as evidence
for viewing noun incorporation as a syntactic process is that the initial noun stem in
a given English nominal compound can't be referred to later in the same sentence,
whereas such can be done with the incorporated noun stem in some polysynthetic
examples of noun incorporation (see Section 3). There are many examples of
English noun incorporation in which the incorporated noun stem can't be
referenced, or at least the sentence becomes lexically ambiguous. For example,
consider the following data:
27
14) *Manta rays filter-feed during the night, and clean them the following morning
(meaning 'manta rays clean their filters the following morning').
15) ?John went berrypicking yesterday, and ate them this morning for breakfast
(meaning 'John ate the berries for breakfast that he picked yesterday').
This raises the question of whether 'noun incorporation' in English is actually
just verbal compounding, lacking the syntactic traits uniquely found in many
examples of noun incorporation, as previously mentioned in Section 3. 1 argue that
the evidence presented by Barrie and Mathieu (discourse reference and adjunct
stranding) consists merely of processes that may occur in some instances of noun
incorporation, rather than universals that every language's noun incorporation
must allow for. Feist likewise foresees this issue and admits, "I simply accept, with
Mithun, Rice & Prideaux, and others, that some constructions that may be called
'compounds' are instances of 'incorporation'" (Feist 2013:162). As previously stated
in this paper, 1 accept Baker's definition of noun incorporation as a syntactic
process, thereby differentiating it from compounding. While 1 don't discuss this
possible wrinkle any further, this is an area begging for future research to be done
(for additional discussion validating noun incorporation in German and English, see
Barrie & Spreng 2009; Barrie 2011; Feist 2013).
5.3 Noun Incorporation in German
Noun incorporation also appears in German, although solely in the progressive
aspect (Barrie & Spreng 2009). Nevertheless, Type 1 certainly does occur in
German, as many verbs in the progressive aspect can become intransitive by
incorporating their object. Additionally, German noun incorporation can only occur
28
when preceded by the verb 'to be' (sein). While this may seem unusual, restricting
the verb stems that can occur in (or in this case, accompany) noun incorporation
appears in other languages as well. In Chamorro, as previously stated, the only verb
stem that can incorporate a noun stem is the Chamorro equivalent of 'to be' (see
Chung and Ladusaw 2004). Returning to German, consider the following example of
Type 1 noun incorporation:
am Wasser trinken.
I be.1SG.PRS PRT water drink.INF
'I am drinking water.'
16) Ich bin
(Barrie & Spreng 2009:375)
Note that the verb stem has incorporated the noun stem, and acts as a single,
unified lexeme. The noun stem in sentence (16), Wasser, could be replaced by any
other 'drinkable' noun stem (e.g. Wein) and remain a grammatical word of German.
While German noun incorporation might only occur in the progressive form, and
thus as a gerund, the compositional process leads to a seemingly lengthy list of
possible derived verb stems. In this regard, German noun incorporation appears to
be quite similar to German nominal compounds.
Barrie and Spreng have made a number of observations regarding the
individual constituents of German incorporation as well. For example, they note that
"nothing can intervene between the verb and the incorporated noun" (Barrie &
Spreng 2009:380-1). Their findings leave the productivity of German Type 1 noun
incorporation closely resembling that of English, in that both have a preference for
the nonfinite form. However, German nominal compounds often have a 'compound
marker,' a grammatical morpheme, inserted in between the two lexical morphemes
ofa noun-noun compound. As Mukai writes of this morpheme, "the inflectional class
29
of the left-hand constituent determines whether a linking morpheme occurs and
what kind" (Mukai 2013:39). For example, consider the following German nominal
compound, in which the linking morpheme -er- (indicating a nominative plural)
appears:
17)Kind-er-wagen.
Child -PL-cart
'Buggy.'
(Mukai 2013:39)
Such is not possible with German incorporated nouns, as the morpheme boundary
cannot be broken, and the noun stem can't be modified. Additionally, German
instances of noun incorporation may only exist as gerunds, and therefore cannot be
conjugated. For example, consider the following sentences:
18)
a) Ich bin
am Bucher verkaufen.
be.1SG.PRS PRT books selL
'I'm selling books.'
b) Ich verkaufe Bucher.
I sell
books.
'I sell books.'
(Barrie & Spreng 2009:377)
Note that in (18), when one attempts to conjugate the incorporated noun,
doing so separates the two constituents and furthermore alters the word order of
the sentence. In this regard, German noun incorporation productivity appears to be
inferior to that of English.
As a result of this initial comparison, two observations can be made. First, the
productivity of noun incorporation within both English and German appears to be
inferior to nominal compounding. This is in part due to the limited verb forms that
can appear in both languages (e.g. the German progressive and English nonfinite
verb stems), as well as the prominence of lexicalized incorporated nouns in English.
30
Additionally, in both languages there are few verb stems (or noun stems) that
appear in more than one instance of noun incorporation. Because of the
lexicalization in English, lexical morphemes can't be freely substituted for one
another, while in German the incorporated nouns (and the incorporating verb
stems) must be preceded by the verb sein ('to be') as well as the pronoun am or bei.
Secondly, English noun incorporation appears to be more productive than
German noun incorporation as per Feist's definition of productivity in noun
incorporation, the opposite of what one might assume given Berg et aJ.'s research
(however, as previously noted, it would be foolish to heavily rely on such an
assumption). While I have shown that Types 1-3 occur in English, I have only
demonstrated that Type 1 occurs in German, and only in the German progressive
form. In order to test the validity of my second observation, it is imperative to elicit
judgment from a native speaker of German.
6
Native German Speakers
For my research, I interviewed two native German speakers in order to determine
whether or not Types 2-4 of noun incorporation can be grammatically constructed
in German. Both German speakers are professors at Swarthmore College, one in the
German department and the other in the philosophy department. The takeaways
from each interview were markedly different in that many examples of German
incorporated nouns were acceptable to one speaker but not the other. While this
may be due to a dialectal difference (one of the speakers was from Northern
Germany, whereas the other grew up in Southern Germany), this contrast between
31
the speakers' lexicons was unexpected. The handout given to each professor was
identical, containing the same questions and example sentences.
The handout presented to the native speakers contained examples of Type 1
and Type 2 noun incorporation, as my primary goal in this experiment was to
determine whether or not Type 2 noun incorporation (transitivity) is possible in
German. In order to make sure that Type 1 noun incorporation is grammatical in
German, I likewise presented both professors with data from Barrie & Spreng 2009,
Barrie 2011, and Feist 2013, as well as asked them to translate English sentences
containing Type 1 incorporated nouns into German. In doing so, I aimed to reinforce
Barrie & Spreng's argument that Type 1 noun incorporation is grammatical in
German (and subsequently, that the process of noun incorporation occurs in
German).
To carry out this study, I presented each professor with two variations of data.
The first set of data was comprised of German sentences containing incorporated
nouns. These German sentences were either taken from pre-existing works or
formed by myself using the sentence structure established in pre-existing works.
For example, consider the following examples, of both Type 1 (19, 20) and Type 2
(21) noun incorporation. English translations have been included, although they
were not presented on the original handouts created for this study:
19)Jch bin
beim Apfel essen.
be.1SG.PRS atthe apple eating
'I am (busy) apple-eating.'
(Barrie & Spreng 2009:378)
20)Der Mann ist beim Wildschweine jagen.
The man is atthe boar.PL
'The man is boar-hunting.'
hunting.
(Barrie 2011:138)
32
21 )John babysittet meine Kinder.
john babysits my kids
'john babysits my children.'
The second dataset consisted of English sentences containing incorporated
nouns. The purpose of these examples was to determine whether or not the same
phrase (containing noun incorporation) would be grammatical upon being
translated to German, as well as to understand how each native speaker would
convey the same message in everyday conversation. For example, consider the
following sentence in English, containing the incorporated noun 'babysit.' While
(22) is the exact English translation of (21), and thus may seem redundant at first
glance, the purpose of this sentence is to differentiate between Type 2 noun
incorporation being a grammatically acceptable feature (21) and a feature that a
native German speaker would actually produce (22). 1 also tested for Type 1 noun
incorporation, as seen in sentence (23):
22)john babysits my children.
23)1 am (busy) apple-eating.
Speaker A, the philosophy professor, was the first native German speaker that 1
interviewed. When presented with sentence (19), in which Type 1 noun
incorporation occurs, he explained that such constructions were colloquiaL
Therefore, although he would understand the meaning of sentence (19), Speaker A
claimed that he would never produce this structure in a conversational context.
When asked how he might convey the concept of himself eating apples, Speaker A
provided the following sentence, in which noun incorporation doesn't occur:
33
24 )lch esse Apfel.
1 eat apples.
'I eat apples.'
Furthermore, Speaker A found some examples of Type 1 noun incorporation to
be ungrammatical, or questionable at best. For example, he remained unconvinced
that one might be able to produce (20) as an acceptable construction in German.
However, Speaker A also provided a number of instances of Type 1 noun
incorporation that he felt to be acceptable in German, including the following words:
Gitarre speilen ('playing guitar'), Vokabular iiben (,practicing vocabulary'), Marathon
laufen ('running a marathon'), and Eis essen ('eating ice cream'). What can be
gleaned from this interview is a confirmation of Barrie & Spreng's argument that
Type 1 noun incorporation exists in German. Speaker A also mentioned that
constructions containing Type 2 noun incorporation, as seen in sentence (21), were
ungrammatical in German, and he couldn't think of any alternative formations of
Type 2 noun incorporation that he would consider grammatical.
Additionally, Speaker A confirmed the observation that German incorporated
nouns may only surface as gerunds, explaining that the conjugated form of Wein
trinken ('drinking wine') necessitates separating the verb stem from the noun stem,
returning the noun stem Wein to the original position of the direct object (following
the verb stem). Finally, Speaker A also noted that certain verb stems in German
appear to be especially productive in their ability to incorporate noun stems, such as
laufen (running), uben ('practicing'), lernen ('learning'), spielen ('playing'), trinken
('drinking') and essen ('eating'). Feist made a similar observation in his analysis of
34
English noun incorporation, as he noted that specific verb stems appear frequently
in English incorporated nouns.
Speaker B's vocabulary appeared to allow for a heightened frequency (and
perhaps productivity) of noun incorporation than seen in Speaker B's dataset. When
presented with the Type 1 incorporated nouns seen in (19) and (20), she recognized
both sentences as acceptable constructions in German, and quickly noted that she
produces similar constructions in a conversational context. In fact, Speaker B
disagreed with the Speaker!\s view of Type 1 noun incorporation as a colloquial
phenomenon, perhaps hinting at a dialectal disparity between Northern and
Southern Germany. Speaker B likewise produced a number of examples of Type 1
noun incorporation from her own vocabulary, such as the following: waschemachen
('doing laundry'), lobsingen ('singing praises'), Steak braten ('frying steak'),
Schlittschuhlaufen ('ice skating'), handverlessen ('handpicking'), and segelfliegen
('hang gliding'). Speaker B was also quick to note, just as Speaker A did, that these
formations may only occur as gerunds.
Speaker B, however, considered my example of Type 2 noun incorporation,
sentence (21), to be a grammatical sentence of German (although she readily
mentioned that she might be incorrect in her remark). If such is the case, the
grammaticality of sentence (21) presents two major challenges to previous work
done on German noun incorporation. First, in sentence (21) the derived verb stem is
not a gerund, and has been conjugated - a characteristic that is seemingly
unacceptable in German noun incorporation. Secondly, in order for sentence (21) to
be grammatically acceptable, it would seem as though Type 2 noun incorporation
35
must exist in German, since the derived verb stem remains transitive. I argue that,
even if we accept the grammaticality of sentence (21), it does not provide evidence
for either the existence of Type 2 noun incorporation in German, or for the possible
conjugation of German incorporated nouns.
In sentence (21), the verb babysittet is an English loanword. Therefore, while the
derived verb stem babysit operates as an example of noun incorporation in English,
it does not do so in German. Consider this word from Baker's syntactic view of noun
incorporation, in which head-fronting occurs in the underlying structure, leading to
a verb stem incorporating a noun stem to produce a newly derived verb stem.
Because babysittet is an English loanword, no such head-fronting occurs in German,
since the underlying form of the loanword has already been head-fronted in English
and subsequently lexicalized. Thus, babysitten does not act as an incorporated noun
in German because it isn't an instance of German noun incorporation - it is a loaned
verb stem that can therefore be conjugated and remain transitive. Furthermore,
neither Speaker A nor Speaker B were able to think of any other constructions
resembling noun incorporation in which the derived verb stem was able to
conjugate and/or remain transitive, providing no further evidence that Type 2 noun
incorporation can appear in German.
In this pilot study, both professors were able to identify sentences such as (19)
as grammatical representations of noun incorporation, and both were additionally
able to produce sentences containing Type 1 incorporated nouns in German by
utilizing the German progressive form. Additionally, neither professor was able to
provide a grammatically acceptable construction of Type 2 noun incorporation in
36
German. Thus, both professors interviewed have, at the most basic level, confirmed
Barrie & Spreng's finding that Type 1 noun incorporation is a grammatical feature of
German, while seemingly no other types may occur.
7
Conclusion
As a result of this pilot study, I can make two observations that are essential in
comparing the noun incorporation processes of English and German. First, I argue
that noun incorporation is more productive in English than in German. Using
Mithun's typology, Types 1-3 of noun incorporation are acceptable in English,
whereas only Type 1 is acceptable in German. In other words, noun incorporation in
German must remain transitive and the incorporated noun stem can't semantically
background previously known or incidental information. Furthermore, Type 1 noun
incorporation is restricted in German so that only one verb stem ('to be') can
precede instances of noun incorporation, whereas no such constraint is placed upon
English. Finally, the derived verb stems in German can only appear as gerunds, and
thus can't be conjugated. While I do not have access to a corpus study comprised of
incorporated nouns in German and English (such as Berg et al. in their discussion of
German and English nominal compounding), I argue that the productivity quantifier
utilized by Berg and his colleagues in their 2012 paper would find English noun
incorporation to be a more productive process than its German counterpart. A
compelling area for future research would be to replicate the procedure of Berg and
his colleagues in order to confirm this paper's findings, as well as to compare the
complexity and frequency of noun incorporation in German and English.
37
The second observation that can be made is that these results shouldn't be
particularly surprising, even considering the work done by Berg and his colleagues.
While Berg et al. (2012) find that German nominal compounding is more productive
and more frequently used in literature than English nominal compounding, one
can't expect noun incorporation to replicate these findings. For, as Baker (2009) and
Barrie and Mathieu (2016) both argue, noun incorporation and lexical compounding
aren't closely related processes despite their surface resemblance, since the former
is syntactic and the latter is largely morphological. A final potential area offuture
research involves interviewing a larger subset of native German speakers in order
to further build on the possible (and impossible) structures of German noun
incorporation as well as to determine if the productivity of German noun
incorporation varies based on the dialect of German being spoken.
38
Appendix A
The following legend consists of morpheme abbreviations that appear in the
interlinear glosses of this paper.
ABS
=
absolutive
AGR = agreement
IMP
imperative
=
INCOMP
=
non-complete verb in Yucatec Mayan
INDIC = indicative
INF
infinitive
=
MIK = specific case marker in Greenlandic Inuktit
PART
=
intransitive participial
PASS
=
passive
PL = plural
PRS
=
present
PRT
=
particle
REL = relative
TR = transitive
UNM
=
unmarked morpheme in Chamorro
39
References
Aronoff, Mark & Nanna Fuhrhop. 2002. Restricting suffix combinations in German
and English: Closing suffixes and the monosuffix constraint. Natural
Language and Linguistic Theory 20.3. 451-490.
Baker, Mark C. 1988. Incorporation: A theory ofgrammatical function changing.
Chicago: University of Chicago Press.
Baker, Mark C. 1996. The polysynthesis parameter. Oxford: The Oxford University
Press.
Baker, Mark C. 2003. Lexical categories. Cambridge: Cambridge University Press.
Baker, Mark C. 2009. Is head movement still needed for noun incorporation? Lingua
119.2.148-165.
Barrie, Michael. 2011. Dynamic asymmetry and the syntax of noun incorporation.
London: Springer.
Barrie, Michael & Bettina Spreng. 2009. Noun incorporation and the progressive in
German. Lingua 119.2. 374-388.
Barrie, Michael & Eric Mathieu. 2016. Noun incorporation and phrasal movement.
Natural Language and Linguistic Theory 34.1. 1-51.
Bauer, Laurie. 2006. Compounds and minor word-formation types. The Handbook of
English Linguistics. Malden: Blackwell Publishing. 483-506.
Berg, Thomas, Sabine Helmer, Marion Neubauer & Arne Lohmann. 2012.
Determinants of the extent of compound use: A contrastive analysis.
Linguistics 50.2.269-303.
Berton, Andre, Pablo Fetter & Peter Regel-Brietzmann. 1996. Compound words in
large-vocabulary German speech recognition systems. Spoken Language
1996.ICSLP 96. 1165-1168.
Chung, Sandra & William A. Ladusaw. 2004. Restriction and Saturation. Cambridge:
The MIT Press.
Feist, Jim. 2013. Noun incorporation in English. Functions of Language 20.2. 159184.
Johns, Alana. 2009. Additional facts about noun incorporation (in Inuktitut). Lingua
119.2.185-198.
Krott, Andrea, Robert Schreuder, R. Harald Baayen & Wolfgang U. Dressler. 2007.
Analogical effects on linking elements in German compound words. Language
and Cognitive Processes 22.1.25-57.
Kunter, Gero. 2011. Compound stress in English: The phonetics and phonology of
prosodic prominence. Berlin: De Gruyter.
Massam, Diane. 2009. Noun incorporation: Essentials and extensions. Language and
Linguistics Compass 3.4.1076-1096.
Meyer, Ralf. 1993. Compound comprehension in isolation and in context: The
contribution of conceptual and discourse knowledge to the comprehension of
German novel noun-noun compounds. Tubingen: De Gruyter.
Mithun, Marianne. 1984. The evolution of noun incorporation. Language 60.4.847894.
40
Mukai, Makiko. 2013. Recursive compounds and linking morpheme. International
Journal of English Linguistics 3.4.36-49.
Plag, Ingo, Gero Kunter, Sabine Lappe & Maria Braun. 2008. The role of semantics,
argument structure, and lexicalization in compound stress assignment in
English. Language 84.4. 760-794.
Rice, Sally & Gary Prideaux. 1991. The case of object incorporation in English.
Proceedings of the 17th annual meeting of the Berkeley Linguistics Society.
Sherkina-Lieber, Marina & Kumiko Murasugi. 2015. Noun incorporation and case in
heritage Inuktitut. Proceedings the 2015 Annual Conference of the Canadian
Linguistic Association.l-ls.
Stvan, Laurel Smith. 2009. Semantic incorporation as an account for some bare
singular count noun uses in English. Lingua 119.2. 314-333.
Weggelaar, C. 1986. Noun incorporation in Dutch. InternationalJournal of American
Linguistics 52.3.301-305.
41